AWS Certified Solutions Architect – Professional — Question 294

A company is planning to migrate an existing high performance computing (HPC) solution to the AWS Cloud. The existing solution consists of a 12-node cluster running Linux with high speed interconnectivity developed on a single rack. A solutions architect needs to optimize the performance of the HPC cluster.
Which combination of steps will meet these requirements? (Choose two.)

Answer options

• A. Deploy instances across at least three Availability Zones.
• B. Deploy Amazon EC2 instances in a placement group.
• C. Use Amazon EC2 instances that support Elastic Fabric Adapter (EFA).
• D. Use Amazon EC2 instances that support burstable performance.
• E. Enable CPU hyperthreading.

Correct answer: B, E

Explanation

Deploying Amazon EC2 instances in a cluster placement group (Option B) is essential for HPC workloads because it places instances physically close together within a single Availability Zone, delivering the low-latency and high-throughput network performance required to mimic a single-rack setup. Enabling CPU hyperthreading (Option E) allows the system to maximize compute throughput for multi-threaded parallel processing tasks common in HPC. Distributing instances across multiple Availability Zones (Option A) or using burstable performance instances (Option D) would introduce latency and inconsistent CPU performance, which degrades HPC efficiency.